Fractures related to osteoporosis are an important cause of morbidity, and in the case of hip fractures, of mortality in the older male population. Since osteoporosis is asymptomatic until a fracture occurs, successful reduction of the societal burden of fractures in men will require that strategies that can identify the subset of men who are at high risk of suffering a fracture, and who should be treated to reduce that risk. Current prediction models have had only partial success achieving this goal, and models incorporating new aspects of bone biology such as bone microstructure are needed to improve fracture prediction in men. In order to be practically applicable to screening a large subset of the overall population as older men represent, such measures need to inexpensive, practical, and easy to implement in clinical practice. Trabecular Bone Score (TBS), a surrogate measure of trabecular microarchitecture, can be easily measured on spine bone density scans (that are routinely done as part of bone density tests) simultaneously as spine bone density is measured. TBS has been shown to be a predictor of major osteoporotic, hip, and clinical vertebral fractures in women, but how well it predicts fractures in men has never been tested. Predictors of changes of TBS over time also have not been explicated. Since TBS is driven by that component of bone (trabecular) with high metabolic activity, there is the possibility that TBS could be a more sensitive indicator of changes in bone strength than areal BMD of the spine (which is often confounded by degenerative changes and aortic calcification) or of the hip (which is mostly cortical bone). Moreover, preliminary evidence from one trial suggests that TBS significantly improves with treatment with at least one therapeutic agent (denosumab), and there is the possibility that TBS could be a more sensitive indicator of response to pharmacologic fracture prevention therapy than other currently available measures. In this ancillary study to the main Osteoporotic Fractures in Men (MrOS) study, and in response to PA-11-261, we propose to measure TBS on the spine bone density scans obtained on 5,994 participants at the baseline visit and 4,485 men attending the second study visit. Our specific aims are; 1) Test the hypotheses that TBS is independently associated with incident hip, radiographic vertebral, and major osteoporotic fractures (clinically evident fractures of the hip, spine, proximal humerus, or distal radius); 2) Test the hypothesis that the addition of TBS to fracture prediction models in older men improves their discrimination of those who will have a fracture from those who will not; and 3) Determine predictors of changes of TBS score between MrOS visits 1 and 2. Establishing the clinical utility of TBS in fracture risk assessment in men will lay the foundation for application in clinical practice to improve fracture prediction and reduce the societal burden of osteoporotic fractures in men.